Resistive & Permanent Magnets for Accelerators

J. G. Weisend IINational Superconducting Cyclotron Lab

Michigan State University

Introduction• Magnets are an important part of accelerator

technology– Many accelerators are “mostly magnets”

• Particle accelerators operate with charged particle beams (e.g. electron, protons, ions, positrons, antiprotons etc)– Interaction of these charged particles and the

magnet field allows manipulation of the beam

2J. G. Weisend II

BveF

Uses of Magnets in Accelerators• Bending or Directing of Beam

– Dipoles

• Focusing of Beam– Quadrupoles

• Beam Correction– Sextupoles and higher

• Beam “Compression” or Energy Separation– Chicanes using dipoles

• Background field for detectors– Solenoids

• Moving charged particles to create synchrotron light– Wigglers, Undulators

A key issue is field uniformityHaving the right field in the right place

Magnetic Design• Modern magnets are designed using a variety of

modeling software (both 2D & 3D)– ANSYS– MAXWELL– TOSCA– POISSSON and PANDRIA– OPERA

• Programs are based on standard EM theory and most use Finite Element Analysis (FEA)

• Programs take into account material properties• Some approximate equations will be given later

Magnet Technology Choices• Permanent• Resistive• Superconducting

Relative MeritsType Advantages Disadvantages

Permanent CompactLow cost ( in small low field magnets)No utilities requiredNo maintenanceSimple to operate – no adjustments requiredCan result in very precise fields (FEL undulators)

Constant field (mostly)Limited in field

Resistive Variable fieldNo need for complicated cryogenic or vacuum systemsCan be built in house or through existing industrial baseRelatively low capital cost

Limited in field (up to ~ 2 T)May require large amounts of electrical power and cooling water – complicatedPossible tritium issuePossible large operating costs for power & water

Relative MeritsType Advantages Disadvantages

Superconducting High fieldVariable fieldLower operating costsReliabilityCold beam tubes yield very high vacuumsCan be made compact

High capital costsLimited industrial baseRequires complicated ancillary systems – cryogenics, vacuum, quench protection

•Technology choice depends on accelerator requirements•All design is compromise

•Requirements for higher energy beams frequently dominate•Permanent and resistive magnets discussed in this talk•Next talk discusses superconducting magnets

Permanent Magnets • Much of this material is taken from Jim Volk

Fermilab Physicist• Permanent magnets are used in very

sophisticated ways in particle accelerators– Fermilab 8 GeV recycler ring is almost entirely

permanent magnets– Precision fields in undulators

• Used where high field strength or field variation is not a requirement– Required field must be well known and unchanging

Permanent Magnet Materials

Note 10,000 Gauss = 1 T CourtesyJ. Volk

Hallbach Array

CourtesyJ. VolkThe easy axis is the direction of field in material

Hallbach Dipole

CourtesyJ. Volk

Hallbach Quadrupole

CourtesyJ. Volk

2 proposals for adjustable field permanent magnets

Hallbach Ring QuadRotating out ring of magnets Changes fieldProposed for NLC

Moveable magneticMaterial adjusts field of quadS. Gottschalk et al.

CourtesyJ. Volk

Effect of Temperature• Field Dependence: Depending on the material, effect

can be significant with moderate temperature change (tens of degree C)

• Even smaller temperature changes can impact alignment & thus function of FEL undulators

• Solutions• Tight environmental controls (LCLS, FLASH)• Temperature compensated design to reduce field

dependence

• Actively cooled devices (s/c & resistive magnets) may have fewer issues - cryostat

ExampleFermilab Recycler Ring

CourtesyJ. Volk

Recycler Design

CourtesyJ. Volk

Recycler Gradient Magnets

CourtesyJ. Volk

Recycler Quads

CourtesyJ. Volk

Note the recycler was ultimately successful

PM Undulators• Moves particle beam back and forth to create synchrotron

light & possibly FEL• Frequently use PM (though resistive & s/c versions also

exist)• Many different types exist ( transverse, helical, with or

without iron)

F. Robles LBNL

PM Undulators• The magnets should be as close to the beam as possible• The precision of the magnetic structure is very important to

success– Requires high mechanical and thermal stability

• Deformation limits of a few micrometers are common• Tends to result in large support structures and tight environmental controls

– Requires precise magnetic measurements & precision alignment

• A wide variety of these devices have been built or proposed

Some Examples

Undulator Design from HASYLabDESY

Undulators for FLASH FELDESY

Cryogenic Permanent Magnet Undulator

• Lowering temperature to 140 K increases field by ~ 13 - 25 % to about 0.5 T(rare earth magnets NdFeB and PrFeB) Radiation damage a concern in NdFeB

• May also help in thermal stability

From “DEVELOPMENT OF A PRFEB CRYOGENIC UNDULATOR AT SOLEIL”

C. Benabderrahmane et al. Proceedings of IPAC’10, Kyoto, Japan

Resistive Magnets• Electromagnets

– Excitation is provided by current carrying conductors ( generally Cu)

– Field is shaped by Iron poles– Field in iron is less than the iron saturation value– Resistive losses in the conductor frequently

require water cooling

• Basis of electromagnets is the Biot-Savart law

R

IB

20

Examples of Ideal Magnets

Dipole

Quad Sextupole

From“Iron Dominated Electromagnets”J. TanabeSLAC-R-754

Some Approximate Equations(assumes iron not saturated)

Dipole Quadrupole Sextupole

Field (B)

Power Dissipated

h

NIB 02

2

02

r

NI

dr

dB

30

2

2 6

r

NI

rd

Bd

0 avgBhjl

P 0

2 )/(2

drdBjlr

P avg0

223 )/(

rdBdjlr

P avg

Where I = current, N = number of turnsh = gap height, r = radius from centerlavg = average length of turnresistivityj = current density = efficiency ( ~ 0.99) = 4 x 10-7

From : Th. ZicklerBasic Magnet DesignCERN Accelerator SchoolBurges – 2009

http://cas.web.cern.ch/cas/Belgium-2009/Bruges-after.html

Comments on Resistive Magnet Cooling

• If the power dissipated via Ohms law is too high than the magnet will over heat and melt

• Thus – the current density must be keep low – yielding

very large conductors or– Cool conductors actively with flowing water

• Flow rates and thus pumping costs can be very high• LCW is generally used (expensive)• Tritium may be produced and have to be dealt with

Fabrication of Resistive Magnets• Proper fabrication of resistive magnets is a very

complicated issue and success requires a great deal of experience

• A very detailed and useful guide here is “Iron Dominated Electromagnets” J. Tanabe, SLAC-R-754 (included in your materials)

• Basically, fabrication can be divided into 3 areas : the yoke, the coil and assembly and installation

Yoke Fabrication• A fundamental requirement is that resistive

magnets should have reproducible performance under excitation & have limited multipole fields

• This requirement is met largely by the iron (steel) yoke – Avoid saturation in the yoke

Yoke Laminations• Yokes are generally made out of laminations to reduce eddy currents thus improving field quality and reducing hysteresis• Laminations are generally shuffled or sorted to allow similar material properties in each magnet• Laminations are carefully shaped to avoid saturation and to deliver the correct field shape•Laminations can be assembled using glue, welding or via mechanical fasteners•For small numbers of magnets with low AC use solid material rather than laminations is frequently used

Yoke AssemblyFrom“Iron Dominated Electromagnets”J. TanabeSLAC-R-754

Coil Fabrication• Wind copper to produce desired field• Insulate via glass tape mylar or other materials

to prevent turn to turn shorts• Install water & electrical connections• Finished coils are typically potted in epoxy to

prevent damage, maintain the coil shape and prevent shorts to ground

Assembly & Installation• The coil, yoke, beam tube and other

assembled together• The magnet fields are carefully measured and

references are installed to allow precise alignment of the magnet with the beam center

• Other tests such as checking water flows and HiPotting ( testing for shorts to ground) are done

Example 1Fermilab Main injector Dipole

• Principle bending magnet in Fermilab Main Injector

• Required to ramp from 0.1 T to 1.73 T at 2.4 T/s

• 344 dipoles were required

Example 1Fermilab Main injector Dipole

From “Design Considerations & Prototype Performance of the Fermilab Main Injector Dipole” D. Harding et al. (PAC 1991)

Example 1Fermilab Main injector Dipole

From “The Design & Manufacture of the Fermilab Main Injector Dipole” B. Brown. (EPAC 1992)

Main Injector and Recycler Magnets in Fermilab Tunnel